1744
VOL.
RUELAS, IRIARTE, KINCL, AND DJERASSI
[CONTRIBUTION FROM
THE
23
RESEARCH LABORATORIES OF SYNTEX, S.A. ]
Steroids. XCVIII.' Synthesis of Some 10~-Hydroxy-19-norsteroids J. PEREZ RUELAS, J. IRIARTE, F. KINCL,
AND
CARL DJERASSI
Received May $9, 1968 The direction and stereochemistry of the acid- and base-catalyzed opening of 5,lO-epoxides of certain 19-norsteroids is discweed and the synthesis of several lOp-hydroxy-19-norsteroidsis reported.
The removal of the angular methyl group at C-10 of certain steroids such as progesterone213 or 17a-ethinyltestosterone4has led to a marked increase in biological activity. This is particularly noteworthy in the latter compound, 19-nor-17aethinyltestosterone (Ib) whose high hormonal activity's4 by the oral route has led to the introduction of this compound (Norlutin) into medical practice. It was felt that it might be of interest to examine the effect of other angular substituents upon biological potency and the present paper is concerned with certain lop-hydroxy-19-norsteroids. Pederson and collaborators6 reported recently that microbiological hydroxylation of 19-nortestosterone led in poor yield to a 10-hydroxy derivative, whose structure was confirmed by osmium tetroxide hydroxylation' of the p, y-unsaturated precursor IIa* of 19-nortestosterone (Ia) followed by dehydration of the intermediate glycol. The stereochemistry of the introduced 10-hydroxyl group was not est,ablishedby the Upjohn group5but conclusive evidence in favor of the 108-orientation could be provideds by noting the coincidenceof the rotatory dispersion curve of 10-hydroxy-19-nortes-
tosterone (IVa) with thatlo of 19-nortestosterone (Ia), where the lop-orientation is established. If the hydroxylation product had been the loa-isomer VIa, then the rotatory dispersion curve would have been of an antipodal type. l1 Consequently, lO&hydroxy-lg-nortestosterone (IVa) can now be employed as the key reference compound for stereochemical considerations in this series. Since we were interested in preparing lobhydroxy-19-norsteroids which might also bear substituents at C-5 (vide infra), the most attractive synthesis of lop-hydroxy-19-norsteroids might well proceed via the 5,lO-epoxide (e.g., 111) of a 5,lOunsaturated 19-nor-3-ketosteroid (11). In fact, earlier work from this laboratory12 had demonstrated the facile conversion of the epoxy ketone A by alkaline treatment to the unsaturated hydroxy ketone B and the structural situation should be completely analogous in a 5,10-epoxy-3-ketone (111). Nevertheless, there exists a patent claim13 that epoxidation of IIa leads to a sharp-melting epoxide (IIIa or Va) which upon exposure to alkali furnishes both C-10 epimeric hydroxy-19nortestosterones (IVa and VIa). The mechanistic -unlikeliness of such a reaction-assuming the (1) Paper XCVII, D. A. McGinty and C. Djerassi, Ann. epoxide to be homogeneous'-prompted us to reN. Y . A d . Sci., 71, 500 (1958). (2) C. Djerassi, L. Miramontes, and G. Rosenkranz, examine the epoxidation of IIa and to establish J . Am. C h .Soc., 75,4440 (1953).
(3) G. W. Barber and M. Ehrenstein, Ann., 603,89 (1957). (4) C. Djeraeai, L. Miramontes, G. Rosenkranz, and F. Sondheimer, J. Am. Chem. Soc., 76, 4092 (1954). (5) R. L. Pederson, J. A. Campbell, J. C. Babcock, S. H. Eppstein, H. C. Murray, A. Weintraub, R. C. Meeks, P. D. Meister, L. M. Reineke, and D. H. Peterson, J. Am. Chem. SOC.,78,1512 (1956). (6) A. J. Birch, J. Chem. Soc., 367 (1950);A. L. Wilds and N. A. Nelson, J . Am. Chem. Soc., 75, 5366 (1953); J. A. Hartman, A. J. Tomasewski, and A. S. Dreiding, J . Am. C h .Soc.. 78,5662 (1956). (7) R. L. Pederson and J. C. Babcock, U.S. Patent 2,806,862. (8) A. J. Birch and S. M. Mukherji, J. Chem. Soc., 2531 (1949). (9) C. Ujerasai, R. Riniker, and B. Riniker, J. Am. Chem. Soe., 78,6377 (1956).
RO
RO
I
H
A
H
B
(10) C. Djerassi, R. Riniker, and B. Riniker,
J. An'.
Chem.Soc., 78,6362 (1956). (11) See C. Djerassi, M. Ehrenstein, and G. W. Barber, Ann., 612,93 (1958). (12) C. Djerasai, 0. Mancera, J. Romo, and G. Rosenkranz, J . Am. C h . Soc., 75, 3505 (1953). (13) F. B. Colton, U.S. Patent 2,729,654. (14) The physical constants of this epoxide are in reasonable agreement with those found in our laboratory for a homogeneousspecimen.
NOVEMBER
1958
STJZROIDS. XCVIII
precisely the stereochemistry of the resulting epoxide and of its transformation products. Treatment of As (lo)- 19-norandrosten-178-01-3one (IIa)8 with monoperphthalic acid a t low temperature furnished in 65% yield a pure epoxide which is assigned the 58,108-stereochemistry (IIIa),since uponheating withmethanolic potassium hydroxide solution it was transformed smoothly ihto the known61s108-hydroxy-19-nortestosterone (IVa). By the same sequence of reactions, the Plyunsaturated isomer IIbl6 of 19-nor-17aethinyltestosterone (Ib)' was converted into the 5/3,108epoxide IIIb and rearranged with alkali to 108hydroxy-19-nor-17a-ethinyltestosterone (Ivb). The 10P-orientation followed from analogy to the course of this reaction sequence in the 19-nortestosterone series (Ia, IIa, IIIa, IVa) and from the fact that its rotatory dispersion curve was nearly identical with thatlo of 19-nortestosterone (Ia). The predominant formation of the 58,108epoxide (111) is noteworthy since alpha attack is usually favored among steroids and production of the a-epoxide (V) might have been expected. However in the absence of the angular methyl gro-up a t C-10, the steric factors controlling a p proach of the reagent are rather subtle and in particular, it should be noted that the 5a,10aepoxide would contain the unfavorable 9,lO-syn backbone in contrast to the 9,lO-anti situation existing in the ,&epoxide I11 which may well represent the controlling factor. As has been reported earlier,'~~19-nor-17aethinyltestosterone (Ib) is an extremely powerful, orally effective progestational agent. In a preliminary Clauberg assay in rabbits, the l0g-hydroxy analog IVb possessed1s only about one-fourth the oral progestational activity of Ib. We can conclude tentatively, therefore, that substitution of the C-10 angular methyl group by hydroxyl does not have the biological potentiating effect of substitution by hydrogen. Introduction of fluorine at various positions of the steroid molecule often results in interesting biological properties" and this applies also to progesterone. It was decided, therefore, to prepare some fluorine-containing 19-norsteroids by the boron trifluoride procedure of Henbest and Wrigley,ls although it was appreciated that fluorohydrins have generally been obtained from trialkylated epoxides while the few tetra-substituted oneslg led to dienes. In view of the fact that the boron trifluoride-promoted opening of epoxides js (15) F. B. Colton, U.S. Patent 2,725,389. (16) The substance exhibited anti-estrogenic activity in mice at a do= of 4007 when employing 0.47 of estrone. (17) See J. Fried and E. F. Sabo, J . Am. C h . Soc., 79, 1130 (1957) and earlier references. (18) P. Tannhauser, R. J. Pratt, and E. V. Jenaen, J . Am. Chem. SOC.,78,2658 (1956). (19) H. B. Henbeat and T. I. Wrigley, J. Chem. Soc., 4596,4765 (1957).
1745
very sensitive to electronic and conformational factorslg no secure a priori prediction about the course of this reaction with 3-keto-5,lOepoxides IIIa and IIIb could be made. When the reaction was performed under the conditions reported in the Experimental section, there was isolated in each case in high yield a single, homogeneous fluorohydrin which on the basis of the usual trans diaxial opening mechanism of epoxidesm could only possess structures VIIa and b or VIIIa and b. Of the two alternatives, VI1 is favored for the reason already advanced above in a discussion of the 58,lop-epoxide formation, namely the presence of an anti 9,10-backbone21in the fluorohydrins VIIa and VIIb. Nevertheless, it was felt that this assignment should be subjected to more secure confirmation and for that purpose, the fluorohydrins VIIa and VIIb were treated with alkali and in each instance yielded the corresponding lop-hydroxy-A4-3-ketones IVa and IVb. This result tends to support formulations VIIa and VIIb by assuming simple base-catalyzed dehydrofluorination, but it is also possible that the reaction proceeas by initial basepromoted ring closurezz of the fluorohydrin back to the epoxide 111. Since either fluorohydrin VI1 or VI11 would yield the same epoxide I11 and since the latter has already been shown above to rearrange to the lop-hydroxy-A*-3-ketoneIV, this alternative course weakens the structure proof of the fluorohydrin. As a result, the epoxy ketone IIIa was treated with perchloric acid in acetone solution to yield a glycol, which was not isolated but which dehydrated directly to the lop-hydroxy unsaturated ketone IVa. The glycol had to possess structure IX or X and since dehydration in this case would not proceed via an epoxide, the orientation of the surviving hydroxyl group of the dehydration product must be identical with that in the glycol. Since the dehydration product was identified as the 108hydroxy derivative IVa, the precursor must have been the 5a,lOj3-glycol IX and we feel justified in assuming the identical stereochemical arrangement for the fluorohydrin (VIIa and b). In connection with the (perchloric) acid opening of the epoxide IIIa, there was also examined the stability of the 108-hydroxy-A4-3-ketomoiety toward acidic reagents since Pederson et aL6mentioned in a preliminary communication that acid-catalyzed dehydration of 108-hydroxy-19-nortestosterone (IVa) led to estradiol without giving any details. (20) A. Furst and P. A. Plattner, 12th Internst. Congress Pure and Appl. Chem. New York, 1951, Abstracts, p. 409. For an example of abnormal epoxide opening see W. S. Knowlea and Q. E. Thompson, J. Am. Chem. SOC.,79, 3212 (1957). (21) The unfavorable truns-syn-trans stereochemistry (aa in VIII) has already been discussed by W. S. Johnson, Expm., 7,315 (1951). (22) Reformation of a fluorohydrin to the epoxide by means of potsssium &butoxidehas already been reported by Henbest and Wrigley (ref. 19).
1746
RUELAS, IRIARTE, KINCL, AND DJERASSI
In our hands, exposure of the 17-acetate of lophydroxy-19-nortestosterone to hydrogen chlorideacetic acid at 5-10' effected the dehydration in nesrly 80% yield with formation of estradiol 17monoacetate (XI). The fluorohydrin VIIb or the epoxy ketone IIIb showed one-fourth or less the oral progestationalza activity of 19-norethinyltestosterone (Ib).ls4 loaHydroxy-19-nortestosterone IVa and the fluorohydrin VIIb were examined for androgenic and anabolic activity** in immature male rats using testosterone as standard and were found to exhibit anabolic-androgenic ratios of 1.5 and 1.7, respectively.
VOL.
23
EXPERIMENTAL%
6 ~ , 1 0 ~ - 0 ~ 1 9 s t u 1 ~ - 1 7 ~ - o 1 - (IIIa). 3 - o nA e solution of 4clO mg. of A6~1~~-19-norandrosten-17~~l-S-one in 10 cc. of chloroform was left with 13.8 cc. of a 0 . W ethereal solution of monoperphthalic acid a t -70" for 2 hr. and then for 18 hr. a t 5-10' whereupon the consumption of reagent corresponded to 0.92 molar equivalents. Dilution with water, extraction with ether, washing with bicarbonate solution and water, followed by drying, evaporation, and recrystallization from acetone-benzene furnished 270 mg. of the epoxido ketone, m.p. 204-205'. The analytical sample was obtained from the same solvent pair and exhibited m.p. 208-210°, [ a ] D -32, A=" 2.74 and 5.83. A d . Calcd. for CI8H,G: C, 74.44; H, 9.03; 0, 16.53. Found: C, 74.32; H, 9.08; 0, 16.47. l O ~ - ~ y d r o x y - 1 9 - ~ ~ s l o s t e s (IVa). o n e The above 519,1019epoxide IIIa (100 mg.) was heated under reflux for 1 hr. with 15 cc. of a 5% methanolic potaaaium hydroxide sohtion, poured into water, extracted with ether and the ether solution waahed well with water and dried. Evaporation and recrystallization from acetone-benzene led to 80 mg. of 10@-hydroxy-19-norteitosterone IVa, m.p. 208-210", [aID -k 80' (methanol), 234-236 m,log a 4.12, A= 3.0 and 6.02 p. Except for some intensity differences" its rotatory dispersion curve (dioxane solution) was identical with that9 oi the microbiological specimen6 (m.p. 199-205", [ a ] f76') ~ and no depression in melting point was observed upon admixture. A n d . Calcd. for C18HzsOs:C, 74.44; H, 9.03; 0, 16.53. Found: C, 73.90; H, 8.87; 0,16.94. Alternatively, 200 mg. of the epoxide 111s in 20 CC. of acetone was left a t room temperature for 16 hr. with 1.5 cc. of 1.5N aqueous perchloric acid and then poured into water. After processing in the usual manner followed by recrystallization from acetone-ether there was isolated 140 mg. of lO@-hydroxy-l9-norteatosterone, m.p. 208-210'. Identity with the above sample waa established by mixture melting point determination and infrared comparison. 6a-Fluoro-10~-h~d~o~-l9-norandrostun-l7~-ol~-one (VIIa). A mixture of 250 mg. of the epoxide IIIa, 30 cc. of dry benzene, 15 cc. of absolute ether, and 0.5 cc. of boron trifluoride etherate was kept a t room temperature for 3 hr., washed with water, dried, and evaporated. Recrystallization of the residue from acetone-benzene furnished 230 mg. of colorless needles, m.p. 203-204', raised to m.p. 215-217' upon repeated recrystallization, [ a ]-41"; ~ no appreciable ultraviolet absorption, A% 2.98 and 5.85 B. A d . Calcd. for CI8HnFO,: C, 69.65; H, 8.77; F, 6.12. Found: C, 70.01; H, 8.59; F, 5.58. When 140 mg. of the fluorohydrin was heated under reflux with 5% methanolic potassium hydroxide solution for 1 hr., there was isolated 90 mg. of lOp-hydroxy-19-nortestosterone (IVa). 66,1O@-Ozido-I 7u-ethiny!-i9-norandrostan-1~ I 9 - o l - h ~ ~ (IIIb). The epoxidation of 400 mg. of 17a-ethinyl-A6(10)19-norandrosten-17~~1-3-0ne (IIb)16 wae performed exactly as described above for IIa and yielded 350 mg. of the epoxide IIIb, m.p. 168-170". Further recrystaJhation from hexane-acetone provided the analytical sample, m.p. 1852.85, 3.05, and 5.85 I(. 187', [ a ] D -75" (methanol), Anal. Calcd. for C&Hl6Ol: C, 76.40; H, 8.34; 0, 15.26. Found: C, 76.63;H,8.36;0,15.24.
AzE
v
d
IV
&
&R
0
v
: F
HO
VI1
'
XI
OH
OH (a)R=H (b)R=CsCH
0
OH
X
(23)These aasays were carried out at the Endocrine Lsborstories, Madison, Wis. (24)h y s by Dr. Ralph I. Dorfman of the Worcester Foundation for Experimental Biology, Shrewubury, Maas.
(25) Melting points are uncorrected. Unless noted otherwise rotations were messured in chloroform solution. All rotation, rotatory dispersion, ultraviolet and infrared meas urements were camed out by Dr. L. Throop and staff. The microanalyses are largely due to Mr. Joseph F. Alicino, Metuchen, N. J. (26) This is probably due to the fact that the product obtained by microbiological hydroxylation' w a ~not completely pure ae judged also by the melting point.
ME METHYL PHENOLIC STEROIDS
NOVEMBER 1958
1747
6 a - F l ~ 1 7 d h i n ~ l - l ~ 1 0 & 1 7 ~ W l 3 a Anal. r C Calcd. for (AH&: C, 76.40;H, 8.34; 0, 15.26. 200 mg. of the Found: C,76.22;H. 8.35;0,15.06. Dehydrath of l O & h y h q ~ - oeetdc l ~ to ~ epoxide IIIb waa carried out a~ described above for IIIa and after recrydalbation from m e t h a n o l - b e n e there d W l 1 7 4 c e f u t e ( X I ) . A current of dry hydrogen chloride waa obtained 160 mg. of the fluorohydrin VIIb, m.p. 247- was pasaed for 2 hr. at 5-10" through a solution of 200 mg. of lO&hydroxy-l%nortestosterone 17-acetaM (m.p. 182249", -39" (methanol). Anal. Calcd. for "FOa: C, 71.83; H, 8.14; F, 5.68. 1S0, [ Q ] D +70°) in 10 cc. of glacii acetic acid. After diluting with water, extracting with ether, washing until Found: C, 71.71; H, 7.99;F,5.47. lQ9-EI&.ozy-17dnyl-l~ (Nb). This neutral, drying, and evaporsting there waa left a adid reaisubstance was obtained in about 80% yield when the due which was recrystallised from acetone-hexane to give epoxide IIIb or the fluorohydrin VIIb was heated under 145 mg. of eetradioi 17-acetate (XI), m.p. 217-218.5". reflux for 1 hr. with 5% methanolic potasgium hydroxide Identity with an authentic specimens waa eetabbhed by solution. The malytical sample cryetdized from acetone mixture melting point determination and infrared comor ethyl acetate and exhibited m.p. 263-264", [ Q ] D 4.5" parison. (methanol), 236 m p , log c 416, 2.95, 3.05, and 6.04u. The rotatow dispersion curve m d in dioxane Am. POSTAL 2679 solution (c, 0.059) was typical^ of a AWkehteroid with MEXICO, D. F. troughsn a t ( ~-556" b and [a]= -665" and a peak at [Qb -604".
m). The boron trifiuoride reaction of
E
+
(27)For nomenclature aee C. Djerasei and W. Klyne, Pm. C k .ik.,55 (1957).
(CONT3IRUTION FBOY TEE
(28) C. Djeraesi, G.Raaenlrrsna, J. Ro", 5.IFsufmsnn, and J. Pataki, J. Am. Chem. rsOe., 72,4534 (1950).
DEPART-NTO F OBSTETRIC8 AND GYNECOLOGY, UmVEESITY OF KANSAS MEDICAL CENTER]
Synthesis of Some 17-Methyl Phenolic Steroids HAROLD J. NICHOLAS Reoeiyed June 9,1958
17a-Methykstrd01was dehydrated with acid to give a product to which the structure 17-methyl-l,3,5(10),l~atehen-3-01 waa amignd The poaition of the double bond was not establhhed unequivocally. Pd-on-charcoal reduction of the tetraene gave two 17-~nethyl-l,3,5(lO)-eatxatriene hmers. Neither of the three new compounds was estrogenidly active at 5-microg"a in preliminarg teating.
in boiling acetic acid, followed by high vacuum sublimation gives rise to a substance, m.p. 157159O, having the structure shown in either IIa or 111. In our hands treatment of 17a-methylestradiol (I) with either hot acetic acid or hydrochloric acid gave a crystalline mixture which could not be resolved by fractional crystallization. Chromatography on Celite-Mg trisilicate afforded a quantitative separation of unreacted I and a second crystalline substance. The latter when p d e d melted at 162-162.5". Julia and Heussefl in a somewhat analogous procedure in the androstane series, ized. 17BdiolThe starting material for their preparation was dehydrated 17-methy1-5-andr&n&p, 17-methyl estradiol, which has been adequately 3-acetate by mild treatment with phosphorus characterjzed and tested;*-* it is about equal in oxychloride-pyridine and (as the diacetate) by estrogenic activity to 17i3-estradiol. It has been treatment with acetic anhydride-pyridine. In reported5that dehydration of 17a-methylestradiol both cases a mixture of 17-methyl and 17-methylene dehydration products were obtained, the latter (1) T.C.Mytua, R. J. Pratt, R L. Morgan, J. O'Donnell, identified by its characteristic methylene absorption in the infrared at 11.36~and 6.03~.Only one and E. V. J-n, J . Am. Chem. &., 77,5655 (1955). (2) Elaevieis - of organic cbmw,series product -ides a small amount of weacted EI, VoL 14, Supplement, Elaevk Publishing Company, starting material) was found on dehydrating 17aMyers et al.' have stated that "it would be of practical as well as theoretical interest if compounds could be discovered which possess little or no primary hormonal activity, but which still have the ability to modify or regulate endocrine balance." With this general objective in mind several 17methyl phenolic steroids of the & m e series were prepared; in preliminary testing the substsnces were estrogenically e t i v e at a 5 microgram level. One of the compounds has been mentioned in the early literature, but was never properly character-
A ? k @ q & a
New York, N. Y., 1956, p. 198%. Several foreign patents are mentioned in this reference. Only the melting point of the free compound ie given for characterization. (3) H. H. Inhoffen and G. Zuhlsdorff, Ber., 74, 604 (1941). (4)B. C. Boe%lsge, H. J. Nicholss, E. A. Doisy, Jr., W. H. Elliott, S. A. Thayer, and E. A. X i , J . Bid. cbm.,202, 27 (1953).
(5) Elaevier's Encgdopdaiz of Organic C h e m w , Series 14, Supplement, Elsevier Publishing Company, New York, N. Y., 1954, p. 1514s. Several foreign patents herein cited give only the m.p. of the free compound. Possible preparation of the compound is mentioned in Ref. 4. (6) 8. A. Julia and H. Heusaer, Heb. Chin. A&, 35, 2080 (1952).
III, Vol.
